A new Consumer Reports investigation released Tuesday found that Uber and Lyft riders often see different prices for the same ride, even when they request it at the same time, along the same route to the same destination.

These price differences may be familiar to Uber and Lyft users who have compared fares with someone else. But the new report found the discrepancies can be dramatic.

Across the rides tested by Consumer Reports, the median gap between the lowest and highest quoted fare for the same trip was a whopping 50%.

The investigation recruited 174 volunteers who priced more than 40 routes on Uber and Lyft across 18 states.

In Kansas City, Missouri, 55 volunteers who checked one particular route generated 29 different prices. In Austin, Texas, the fares for another route ranged from $25 to $65, a 160% difference. Rides in the investigation were requested within minutes of one another, and in many cases within the same minute.

Consumer Reports said some of the routes ordered by volunteers in the study were from exactly the same pinpoint location, but others differed slightly.

Ride-hailing companies have long used dynamic pricing, also known as “surge pricing”, which calculates prices based on shifting market conditions such as rider demand, car availability, and traffic conditions.

But the differences that Consumer Reports found among seemingly identical ride requests raise new questions about whether Uber and Lyft are using personalized pricing or “surveillance pricing.” These systems set a unique price for each individual customer based not only on market conditions, but also on that person’s user data.

Derek Kravitz, the lead author of the report, says Uber and Lyft have access to a huge amount of customer data within the app, and that this information could be used by the companies to predict what a customer might be willing to pay.

“We found that they have the capabilities to look at a lot of different things, namely how you interact with the app, how accurately or fast you type an address… whether or not you’re going to a daycare to pick up a child,” said Kravitz, who analyzed the ride-hailing apps’ privacy policies and patent filings.

Consumer Reports could not determine whether any of these specific factors were actually used to calculate fares, however.

Both Uber and Lyft have denied using surveillance pricing in the past. Reached for this story, an Uber spokesperson told NBC News the company “does not personalize prices, period,” and that differences were “due to changing real-time marketplace conditions.”

The company called the Consumer Reports investigation “flawed,” saying prices can change second-by-second based on demand, driver availability, traffic and even slight differences in GPS location.

Lyft did not reply to a request for comment from NBC News.

The question of whether ride-hailing companies use personal data to determine the maximum price a rider is willing to pay is not a new one.

In 2022, an academic study published in the University of Chicago’s Journal of Law and Economics found that Uber charged riders higher prices for trips to more expensive hotels. It also found, however, that “allowing ride-sharing companies to price discriminate improves travelers’ welfare, on average, by increasing their travel options.”

In March, the Republican-controlled House Committee on Oversight and Government Reform announced an investigation into whether Uber, Lyft, and other companies were using surveillance pricing. Rep. James Comer, Ky., who chairs the committee, warned that pricing algorithms could “weaponize personal data and pad [companies’] profit margins at the expense of providing transparency to consumers.”

At the time, Uber said it did not engage in surveillance pricing, “including in the ways the letter describes,” according to the travel technology news site Phocuswire.

Many factors go into ride pricing, making it nearly impossible to know exactly what is driving any individual price.

Keith Chen, a professor of Behavioral Economics at UCLA and the former head of economic research at Uber, said ride prices on the app have moved toward “more complicated optimization” over time.

Pricing, he said, evolved from a taxi-like model based on time and mileage to more complex factors that include future car demand and randomized pricing tests to gauge what customers were willing to pay.

Chen, who is widely credited with having designed Uber’s surge pricing system, said that during his time at the company he did not see evidence that customer data was being used to determine base prices. But it did play a role in determining discounts.

Both Uber and Lyft have publicly acknowledged that user data can be used to determine discounts. Personalized promotions are a more widely used business practice, such as when a retailer offers a discount to a new customer or sends a coupon to someone who left items in an online shopping cart.

For riders who want to make sure they’re getting the lowest fare possible, experts recommend comparing prices across multiple ride-hailing apps before booking a ride. A recent National Bureau of Economic Research study found that only 1 in 6 riders checks both Lyft and Uber, and prices differ by an average of 14%.